Conventional electron tubes for generating microwaves, such as the traveling wave tube (TWT) and the klystron rely on axial motion of a beam of electrons interacting with axial components of the electric field of a wave-supporting structure. In the TWT the wave velocity must be equal to the electron velocity, so a periodic "slow wave" circuit must be used. For very high frequencies such as millimeter waves, the periodic pitch of the circuit becomes very small, thus hard to fabricate and capable of handling only low power. Also, the circuit diameter must be small compared to a wavelength, and must be close to the beam so that its usful fringing field can interact with the beam.
In the search for higher power at higher frequencies, several "fast wave" tubes have been proposed in which a non-periodic circuit such as a smooth waveguide is used to interact with periodic modulation of the electron beam. In a smooth hollow waveguide, of course, the axial phase velocity of the wave is always greater than the velocity of light so that the beam's axial velocity can never be synchronous with it. A two-conductor line in which the velocity is exactly equal to the velocity of light is also classed as a "fast wave" circuit. An electron would have to have infinite energy to be synchronous with it.
The most successful fast wave tube has been the "gyrotron" in which electrons in a beam are given spiraling cyclotron motions in an axial magnetic field. The electrons become bunched into certain phases of their cyclotron orbits by interacting with a transverse electric field in a smooth waveguide carrying a wave at or near its lower cutoff frequency. The gyrotron has been successful as an oscillator for extremely high power. It will be shown later that its bandwith is inherently small, so it would not be very useful as an amplifier for communications or the like.
Another tube employing cyclotron motion of electrons in a transverse field is described in U.S. Pat. No. 3,183,399 issued May 11, 1965 to Richard H. Pantell and assigned to the assignee of this application. In Pantell's tube a rectangular smooth waveguide is used, supporting a linearly polarized TE.sub.01 wave. Pantell described the beam modulation as due to axial bunching of electrons into a spiral ribbon by velocities induced by the cyclotron motion cutting transverse magnetic field lines of the radio-frequency wave mode. Such bunching certainly may exist, although it now appears that Panell's tube probably operated with gyrotron bunching utilizing slightly relativistic electron motion. Pantell's tube was thus an early gyrotron, and would have a very narrow bandwidth. U.S. Pat. No. 3,249,792 issued May 3, 1966 to Richard H. Pantell describes a variation of the above-described tube which uses a two-wire transmission line instead of a hollow waveguide. The wave velocity is then just the speed of light for all frequencies. FIG. 3 of the latter Pantell patent is an omega-beta diagram from which it is clear that synchronous interaction can occur only at sharply limited frequencies.